CardioLens

• 

  ML Prediction

• 

  Choose Your Mode

• 

  Patient Data

• 

  Welcome Page

• 

  3D Visualization

• 

  Guided Anatomy Tour

Inspiration

• The growing concern around heart diseases inspired us to explore how technology could make cardiac risk information easier to understand.
• We wanted to combine machine learning and 3D visualization to present cardiac insights in a more interactive way.
• This idea led us to build CardioLens — ML-Powered Cardiac Intelligence.
• CardioLens is our team's first properly planned and implemented project, making it a significant milestone for Team Nexus.

What it does

• Cardiac Risk Prediction: Uses Logistic Regression and Random Forest for cardiac risk classification.
• Vessel-Specific Analysis: Focuses on three major coronary arteries: LAD, LCX, and RCA.
• Interactive 3D Visualization: Enables users to explore a 3D heart model.
• Guided Anatomy Tour: Provides a structured visual exploration of heart anatomy.
• Risk Factor Exploration: Presents relevant cardiac risk factors through a dedicated interface.
• Model Interpretability: Explores SHAP-based feature explanations to help understand model predictions.

How we built it

• Frontend: React for the user interface.
• Backend: Python and FastAPI for backend development.
• Machine Learning: Scikit-learn with Logistic Regression and Random Forest.
• Data Processing: Pandas and NumPy.
• Model Interpretability: SHAP for exploring feature contributions.
• 3D Visualization: Three.js-based technologies for interactive heart visualization.
• We divided responsibilities across frontend, backend, ML, and project coordination to develop the prototype collaboratively.

Challenges we ran into

• Preparing and processing the dataset for machine learning.
• Evaluating different models and understanding their performance.
• Representing heart anatomy through interactive 3D visualization.
• Exploring meaningful explanations for model predictions.
• Coordinating different components and managing time within the hackathon deadline.

Accomplishments that we're proud of

• Developed our team's first properly planned and implemented project.
• Built a dedicated frontend for cardiac risk exploration.
• Implemented two machine learning approaches: Logistic Regression and Random Forest.
• Created an interactive 3D heart visualization and guided anatomy tour.
• Explored SHAP-based model interpretability.
• Completed our project presentation and created our Devpost project page.

What we learned

• Practical experience in data processing and machine learning.
• The importance of model evaluation and explainability.
• How 3D visualization can support interactive exploration.
• Better project planning, task distribution, and teamwork.
• The importance of coordinating different software components during development.

What's next for CardioLens

• Complete the remaining backend and ML features.
• Integrate the frontend and backend and test the complete workflow.
• Push the final code to GitHub.
• Deploy the application.
• Record the project demonstration video.
• Improve the 3D visualization and model explanation experience.

CardioLens — Your Heart Through a Smarter Lens.

Built With

• 3d-visualization
• css
• data
• data-science
• explainable-ai
• fastapi
• healthcare
• html
• javascript
• logistic-regression
• machine-learning
• medical-technology
• numpy
• pandas
• python
• random-forest
• react
• scikit-learn
• shap
• three.js